def test_selection_transform(df):
selector = ColumnSelector(["x", "y"])
op = SelectionOp(selector)
result_df = op.transform(ColumnSelector(), df)
assert (result_df.columns == ["x", "y"]).all()
def test_selection_output_column_names(df):
selector = ColumnSelector(["x", "y"])
op = SelectionOp(selector)
result_selector = op.output_column_names(ColumnSelector())
assert result_selector.names == ["x", "y"]
def test_selection_output_schema(df):
selector = ColumnSelector(["x", "y"])
schema = Schema([ColumnSchema(col) for col in df.columns])
op = SelectionOp(selector)
result_schema = op.compute_output_schema(schema, ColumnSelector())
assert result_schema.column_names == ["x", "y"]
def test_workflow_select_by_tags(op):
schema1 = ColumnSchema("col1", tags=["b", "c", "d"])
schema2 = ColumnSchema("col2", tags=["c", "d"])
schema3 = ColumnSchema("col3", tags=["d"])
schema = Schema([schema1, schema2, schema3])
cont_features = ColumnSelector(tags=["c"]) >> op
workflow = Workflow(cont_features)
workflow.fit_schema(schema)
output_cols = op.output_column_names(ColumnSelector(["col1", "col2"]))
assert len(workflow.output_schema.column_names) == len(output_cols.names)
def test_fit_schema_works_with_node_dependencies():
schema = Schema(["x", "y", "cost"])
cont_features = ColumnSelector(["cost"]) >> ops.Rename(postfix="_renamed")
cat_features = ColumnSelector(["x", "y"
]) >> ops.TargetEncoding(cont_features)
workflow1 = Workflow(cat_features)
workflow1.fit_schema(schema)
assert workflow1.output_schema.column_names == [
"TE_x_cost_renamed", "TE_y_cost_renamed"
]
def test_fit_schema_works_with_addition_nodes():
schema = Schema(["x", "y", "id"])
x_node = ColumnSelector(["x"]) >> ops.Rename(postfix="_renamed")
workflow = Workflow(x_node + "y")
workflow.fit_schema(schema)
assert workflow.output_schema.column_names == ["x_renamed", "y"]
x_node = ColumnSelector(["x"]) >> ops.Rename(postfix="_renamed")
y_node = ColumnSelector(["y"]) >> ops.Rename(postfix="_renamed")
workflow = Workflow(x_node + y_node)
workflow.fit_schema(schema)
assert workflow.output_schema.column_names == ["x_renamed", "y_renamed"]
def test_fit_schema_works_with_raw_column_dependencies():
schema = Schema(["x", "y", "cost"])
cat_features = ColumnSelector(["x", "y"]) >> ops.TargetEncoding("cost")
workflow = Workflow(cat_features)
workflow.fit_schema(schema)
assert workflow.output_schema.column_names == ["TE_x_cost", "TE_y_cost"]
def test_fit_schema_works_when_subtracting_column_names():
schema = Schema(["x", "y", "id"])
cont_features = (ColumnSelector(
["x", "y"]) >> ops.FillMissing() >> ops.Clip(min_value=0) >> ops.LogOp
>> ops.Normalize() >> ops.Rename(postfix="_renamed"))
workflow1 = Workflow(cont_features - "y_renamed")
workflow1.fit_schema(schema)
assert workflow1.output_schema.column_names == ["x_renamed"]
def test_fit_schema_works_with_grouped_node_inputs():
schema = Schema(["x", "y", "cost"])
cat_features = ColumnSelector(["x", "y",
("x", "y")]) >> ops.TargetEncoding("cost")
workflow1 = Workflow(cat_features)
workflow1.fit_schema(schema)
assert sorted(workflow1.output_schema.column_names) == sorted(
["TE_x_cost", "TE_y_cost", "TE_x_y_cost"])
def test_fit_schema():
schema = Schema(["x", "y", "id"])
cont_features = (ColumnSelector(schema.column_names) >> ops.FillMissing()
>> ops.Clip(min_value=0) >> ops.LogOp >> ops.Normalize()
>> ops.Rename(postfix="_renamed"))
workflow = Workflow(cont_features)
workflow.fit_schema(schema)
assert workflow.output_schema.column_names == [
"x_renamed", "y_renamed", "id_renamed"
]